german apprenticeship training and the skill weights …00000000-5b2d-ab7b-0000... · german...
TRANSCRIPT
German Apprenticeship Training and theSkill Weights Approach – An Empirical Analysis
Uschi Backes-Gellner, Johannes Mure, Regula Geel
University of ZurichInstitute for Strategy and Business EconomicsPlattenstrasse 14, CH-8032 Zurich,Tel.: + 41 44 634 42 81, Fax: +41 44 634 43 70 [email protected]
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 2
I. The focus of our paper
- Mobility after Apprenticeship Training
- Firm`s Investment in Apprenticeship Training
II. The Skill-Weights Approach (Lazear 2004)
III. Data
IV. Empirical Results
V. Conclusion
Outline
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 3
II. The Skill-Weights Approach (Lazear 2004)
Output of the employee at firm i: BAy iii )1( λλ −+=
• All skills are of general nature.• These general skills are combined differently in different firms .• A, B = General Skills A and B• λi = Relative Weight of the skills A and B in firm i.• yi = Output at firm i.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 4
• Starting Point: The employee has to decide to what extent he acquiresthe skills A and B in order to maximize his income.
II. The Skill-Weights Approach (Lazear 2004)
Output of the employee at firm i: BAy iii )1( λλ −+=
• All skills are of general nature.• These general skills are combined differently in different firms .• A, B = General Skills A and B• λi = Relative Weight of the skills A and B in firm i.• yi = Output at firm i.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 5
Output of the employee at firm i
“ outside (expected)
BAy iii )1( λλ −+=(with A>B)
BAy j )1( λλ −+=
λλ −iLevel of specificity
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 6
Output of the employee at firm i
“ outside (expected)
BAy iii )1( λλ −+=(with A>B)
BAy j )1( λλ −+=
λλ −iLevel of specificity
Worker Mobility
H1 The more specific the qualification requirements of an occupation(Ausbildungsberuf), the lower is the probability of an occupationalchange after apprenticeship training.
Firm`s Training Investment Share
H2 The more specific the qualification requirements of an occupation, thelarger is the investment of the firm in apprenticeship training.
Hypotheses
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 7
BIBB/IAB - Qualification and Career Survey 1979, 1991/92 und 1998/99
Characteristics
• Representative random samples of the German workforce.• Detailed information about skill profiles of the employees.• Information about the vocational carreres of the employees.• Information about training activities of the employees.
Selection
• Persons aged between 16 and 65 years• No civil servants• No self-employed
III. Data
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 8
Source: BIBB/IAB-Survey 1979, own calculations.
IV. Explanatory Variable: Specificity Degree
Qualification Portfolio of a Banker
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 9
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
Source: BIBB/IAB-Survey 1979, own calculations.
Qualification Portfolio of a Banker
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 10
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
Source: BIBB/IAB-Survey 1979, own calculations.
Qualification Portfolio of a Banker
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 11
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
Source: BIBB/IAB-Survey 1979, own calculations.
Qualification Portfolio of a Banker
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 12
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
Source: BIBB/IAB-Survey 1979, own calculations.
Qualification Portfolio of a Banker
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 13
German Qualification Portfolio
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
Qualifications
Source: BIBB/IAB-Survey 1979, own calculations.
Qualification Portfolio of a Banker
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 14
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
QualificationsGermany Banker
Source: BIBB/IAB-Survey 1979, own calculations.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 15
IV. Explanatory Variable: Specificity Degree
0
1
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Relative Frequency
Importance of Qualifications
QualificationsGermany Banker
0
5
10
15
20
25
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Rank
Rank of Qualifications
QualificationsGermany Banker
Source: BIBB/IAB-Survey 1979, own calculations.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 16
Determinants of Mobility (during working life)
V. Specificity of Occupation and Mobility
Specificity of Occupation -.0019132 **(.0007577)
male .001205(.0069017)
age .0395193 ***(.0019426)
age2 -.0003772 ***(.0000242)
vocational training .0869616 ***(.0066178)
further controls
N 25.973Wald chi2(14) 1766.39Prob > chi2 0.0000Pseudo R2 0.0508Log pseudolikelihood -16794.486
Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Datasets 1979, 1991/92 und 1998/99. Further controls:groesselehrbetrieb49, groesselehrbetrieb99, groesselehrbetrieb999, groesselehrbetrieb1000, gemeindegroesse100, gemeindegroesse500, gemeindegroesseueb500, welle79, welle91. Level of significance ***=1%, **=5%, *=10%. All estimates of the probit-Regression are estimatedwith robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 17
Determinants of Mobility (directly after completion of the apprenticeship training)
Specificity of Occupation -.0007744 ***(.000267)
male.0052922 **(.0024538)
age.0006948 (.0006954)
age2-.0000152 *(8.89e-06)
vocational training.0103447 ***(.0026047)
further controls
N 25.973Wald chi2(14) 681.06Prob > chi2 0.0000Pseudo R2 0.0717Log pseudolikelihood -4470.8469
V. Specificity of Occupation and Mobility
Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Datasets 1979, 1991/92 und 1998/99. Further controls:groesselehrbetrieb49, groesselehrbetrieb99, groesselehrbetrieb999, groesselehrbetrieb1000, gemeindegroesse100, gemeindegroesse500, gemeindegroesseueb500, welle79, welle91. Level of significance ***=1%, **=5%, *=10%. All estimates of the probit-Regression are estimatedwith robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 18
Merge of Six Datasets
V. Specificity of Occupation and Firm`s Training Investment
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 19
Determinants of Firm`s Net Investment Cost
V. Specificity of Occupation and Firm`s Training Investment
Specificity of Occupation397.2***(119.6)
variance in the skill portfolio 397.3(2592.0)
age at end of apprenticeship training 1967.8***(569.5)
Const. -33 866.3***(12 279.6)
Number of observations 74
F(3, 70) 5.54
Prob > F 0.00
R2 0.19
Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Datasets 1979, 1991/92 und 1998/99 and the BIBB-cost and benefitsstudies of apprenticeship training 1980, 1991and 1998. Level of significance ***=1%, **=5%, *=10%. All estimates of the OLS-regression areestimated with robust standard errors in order to avoide heteroskedasty.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 20
Transfer to Further Vocational Training
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 21
Transfer to Further Vocational Training
H1` The more unusual the qualification requirements of an employeein his current firm, the lower is the probability of a job change.
H2` The more unusual the qualification requirements of an employeein his current firm, the larger is the training investment share thefirm has to bear.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 22
V. Specificity and Change of Industry
Specificity of Industry-0.0114***(0.0041)
blue collar0.0166(0.0126)
age0.0516***(0.0075)
age2-0.0005***(0.0001)
exp-0.0032(0.0037)
exp20.0001(0.0001)
male-0.0038(0.0129)
N 12 036
Loglikelihood -7512.2049
Prob> X2 ***
Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Dataset 1998/99. Further controls: level of schooling, level of vocational qualification,income, further characteristics if the employer, further characteristics of the employee. Level of significance ***=1%, **=5%, *=10%. All estimates of theprobit-Regression are estimated with robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 23
V. Specificity and Change of Industry (Panel data from the GSOEP)
Specificity of Industry-0.0089***(0.0033)
blue collar0.0134(0.0340)
age0.0010(0.0180)
age2-0.0006**(0.0003)
exp-0.0464***(0.0079)
exp20.0012***0.(0002)
male0.0063(0.0341)
N
Groups48 02712 750
Loglikelihood -10381.053
Prob> X2 ***
Source: Own calculation with STATA 9.2 on the basis of the GSOEP 1996-2003. Further controls: level of schooling, level of vocational qualification,income, further characteristics if the employer, further characteristics of the employee. Level of significance ***=1%, **=5%, *=10%. All estimates of theprobit-Regression are estimated with robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 24
V. Individal Specificity and Firm`s Training Investment Share
Indiv. Specificity (Industry) 1.2337***(0.1574)
Indiv. Specificity (Germany) 1.0843***(0.1198)
Distrubution of λ
Lambda 0.0387**(0.0161)
0.0178(0.0127)
Layoff-risk
RiskWiLage 0.0544(0.0426)
0.0567(0.0420)
Market thickness
AMDreg -0.0114***(0.0033)
-0.0115***(0.0033)
AMDbr -0.0045***(0.0009)
-0.0076***(0.0020)
N 13 738 13 738
Loglikelihood
-17881.24
Prob > X2 ***
-17864.13
Prob > X2 ***
Source: Own calculation with STATA 9.2 on the basis of the BIBB/IAB Dataset 1998/99. Further controls: level of schooling, level of vocational qualification,income, further characteristics if the employer, of the job, of the employee and of the labor market. Level of significance ***=1%, **=5%, *=10%.All estimates of the ZIP-Regression are estimated with robust standard errors in order to avoide heteroskedasty. All coefficients represent marginal effects.
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 25
1. The Skill Weights Approach widens the theoretical analysis of importantquestions related to vocational training.
2. The Skill-Weights Approach is very helpful to explain different empiricalfindings consistently:
- The more specific an investment in human capital, the bigger isthe educational investment a firm has to bear.
- The more specific an investment in human capital, the more reduced isan employees mobility after the training.
- These results are robust when estimating on occupational as well as onindustry level.
- The results hold on the level of initial vocational training as well as on thelevel of further vocational training.
VI. Conclusion
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 26
3. The apprenticeship training in Germany -formerly seen as general training-is very heterogenous in specificity: Some apprenticeships are more generalwhereas others are highly specific compared to the German labor market.
VI. Conclusion
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 27
3. The apprenticeship training in Germany -formerly seen as general training-is very heteogenous in specificity: Some apprenticeships are more generalwhereas others are highly specific compared to the German labor market.
4. Obviously there is a trade-off between firm`s incentive to invest and thespecificity of an apprenticeship training. The higher the specificity,the higher is the firm`s incentive to invest.
VI. Conclusion
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 28
3. The apprenticeship training in Germany -formerly seen as general training-is very heteogenous in specificity: Some apprenticeships are more generalwhereas others are highly specific compared to the German labor market.
4. Obviously there is a trade-off between firm`s incentive to invest and thespecificity of an apprenticeship training. The higher the specificity,the higher is the firm`s incentive to invest.
If somebody has the idea to reduce specificity of theapprenticeship training in order to prepare the employeesfor technological changes, mobility might increase. But, at thesame time, the firms would reduce their training investments.
VI. Conclusion
Backes-Gellner/Mure/Geel. International Conference on Economics of Education. Zurich. 08.07.2008 / 29
Thank you for your attention!
www.economics-of-education.ch